Gearboxes? Help!

[Akash Rastogi]

Though COTS gearboxes are recommended for teams who have not designed or prototyped custom gearboxes in the offseason, there are plenty of resources to help teams design their first custom gearbox, if you would like to give it a try.

To determine PD, OD, and center to center spacing of a gear, WCP has a neat feature on this page: http://wcproducts.net/how-to-gears/

To determine your desired gear ratio, use JVN's mechanical design calculator: http://www.chiefdelphi.com/media/papers/2059

In order to figure out mounting holes for a shaft encoder, check out http://www.chiefdelphi.com/media/papers/2681

Many of these resources make it easy to design a custom gearbox in CAD, but you do have to know what type of speed and ratios you want.

[pfreivald]

The Banebots p80 Gearbox is a workhorse. We've used 12:1 and 9:1 on our drive trains for direct-drive since we abandoned chain in 2008, and a 12:1 on our robot lift (with a 4:1 lead screw) in 2010, and have never had a failure.

The major caution there would be that BaneBots tends to run out very quickly during the build season, so buy early if you're planning on them!

[ToddF]

We are using our summer to experiment with custom gearboxes and building our first west coast style drivetrain.

Here are some pics of our first custom gearbox, toughbox innards with custom sideplates:




We currently have on order a frame and gearbox plates from Team221.com: http://team221.com/viewproduct.php?id=108
When the plates arrive, we plan to machine lightening features to remove even more weight.

We are finding that the custom gearbox route is an interesting balance between cost, machining time, and design flexibility.

[Jon Stratis]

Just wanted to add a personal story about grease in a gear box...

A number of years ago we were at competition and were picked by another team for eliminations before lunch on Saturday. Naturally, we were all excited when we went to eat... then someone from the other team found us and said "we need help".

One of our mentors spent the entire lunch hour rebuilding BOTH of their toughboxes, which had failed in their last match. When it came time to add grease, he asked the team for some. One of their students pulled out the grease tube, and proceeded to squirt some on to a Q-tip. No joke. Apparently, when they put together their old gearboxes, this was how they greased them up. Our mentor asked for the tube, and right in front of their horrified faces squirted half of it into the gearbox.


That said, it is possible to put too much grease into a gearbox. For example, a sealed planetary like you can get from BaneBots will likely have issues if you pack it completely full of grease. However, we've found that the toughbox and CIMple box both have enough places where gear will work its way out that it's not really a problem. When our drivetrain first goes together each year, we let it run for an hour or two with no load, just to help wear in the parts a bit (we find it helps stretch the chain and then it meshes better with the sprockets). We fully expect grease to work its way out of the gearbox when we do this! If it doesn't, then we didn't add enough.

[Tom Line]

Originally Posted by Trent B

I cannot find the pictures online right now (when I do I need to save them because I always have this problem) but your pictures are a deadringer for the wear a pair of our tough boxes showed in 09 as a result of the assemblers not putting in any grease.

What type of grease are you using?

EDIT: Here is the appearance of a completely messed up gearbox, no grease for an entire build season.



Set of photos

Bingo. We had both our toughboxes fail in 2010 because the original grease was cheap garbage. After nearly 100 matches it had turned solid and metal was the only lubricant that we had.

We now regrease before each competition. It's easy because there are holes that allow you to do so on the gearbox without opening it up! Just fill the sucker and tape over the holes.

[AlDee]

Originally Posted by SuperNerd256

I second this. I've looked through multiple gearboxes for use on our robots, and through all of them, the CIMple Box is by far one of the best. Very reliable, easy to mount, and saves teams a lot of money/time when designing a robot (no need to order different gearboxes, these come in the kit of parts!). 1 per side, and your drivetrain will be trucking all year long.

There's a reason that 1114 used CIMple Boxes in their Kitbot on Steroids. They're amazing.

I'm Glad I found this thread. I'm looking to experiment with a transmission that can take two motors. I thought of the CIMple box, because it comes in the KOP, and we have a bunch of them laying around. I was wondering if these were adequate, or was there an advantage to using something like a tough box, etc. From what I see here, the CIMple box sounds fine. Does anyone have alternitive views on a better choice for a dual motor transmission?

[EricH]

Originally Posted by AlDee

I'm Glad I found this thread. I'm looking to experiment with a transmission that can take two motors. I thought of the CIMple box, because it comes in the KOP, and we have a bunch of them laying around. I was wondering if these were adequate, or was there an advantage to using something like a tough box, etc. From what I see here, the CIMple box sounds fine. Does anyone have alternitive views on a better choice for a dual motor transmission?

Define "better".

Single speed or multiple speed?
CIM or other motor?
Durability criteria?
Weight?
What does it need to attach to and how?


Any multi-motor gearbox on AndyMark will pass just about any FRC-type durability test. Most run with CIMs, but some run with others, or adapt other motors to mount like CIMs. Some are single speed, some are multiple speed, some can be stacked to get a higher gear ratio.

[JVN]

Originally Posted by AlDee

I'm Glad I found this thread. I'm looking to experiment with a transmission that can take two motors. I thought of the CIMple box, because it comes in the KOP, and we have a bunch of them laying around. I was wondering if these were adequate, or was there an advantage to using something like a tough box, etc. From what I see here, the CIMple box sounds fine. Does anyone have alternitive views on a better choice for a dual motor transmission?

It all comes down to gearing. The CIMple box has some advantages because it consists of only a single stage of gear reduction, then a single chain reduction to the wheel. Less stages = higher efficiency = less parts = less complexity = less weight.

However, this means the internal ratio is limited and it might not fit in some drive setups.

Have you learned about drivetrain gearing or some of the other considerations of drivetrain design?

What wheels do you want to use? Why?
What configuration do you want to use for your gearbox? (direct drive? gearbox + chain?) Why?
Do you want to push? Why?
Do you want to have a high top-speed? Why?
Do you want 1-speed or 2-speed? Why?

Once you understand some of the tradeoffs there and have a better idea of what gear reduction you need before your wheels, you can then make better decisions about whether the CIMple box works for your application or whether you need something with more reduction, or 2-speeds, or whatever...

If you want to email me, we can talk through the goals you have for your drivetrain, and I can help you understand some of the tradeoffs you'll need to make. Then we can talk about COTS/Custom options to achieve these goals.

john.vneun@gmail.com
Tough to have that back-and-forth on a forum.

-John

[Billfred]

Originally Posted by JVN

What wheels do you want to use? Why?
What configuration do you want to use for your gearbox? (direct drive? gearbox + chain?) Why?
Do you want to push? Why?
Do you want to have a high top-speed? Why?
Do you want 1-speed or 2-speed? Why?

You missed one (arguably) relevant question: How long does it have to run? If it only has to run ten seconds at the end of the match, that's one thing. If it has to run all two minutes, that's another thing. If it has to run all day at a demo, that's yet another thing. (We learned this nearly the hard way; after the season we stripped two CIMs off of our 2011 robot and used it for driver practice. Let's say we didn't get all the runtime we wanted before the motors were too hot to touch.)

[Richard Wallace]

Originally Posted by Billfred

... If it has to run all day at a demo, that's yet another thing. (We learned this nearly the hard way; after the season we stripped two CIMs off of our 2011 robot and used it for driver practice. Let's say we didn't get all the runtime we wanted before the motors were too hot to touch.)

Excellent point, Billfred. Al related a similar story in another thread recently; I recall that in Wildstang's case, they knew they would damage CIMs by continuous driving practice, pausing just long enough to swap in fresh batteries -- and they did it anyway because the practice time was more valuable to them than the CIMs.

Note also that a hot CIM case indicates a VERY hot CIM armature (that rotating electrical coil on the inside). Rated service life for a CIM, per its data sheet, is 1000 cycles of 6 minutes ON followed by 30 minutes OFF; the specified mechanical load during ON time is 4 lbf-in (64 oz-in) at 4320 RPM, with the motor drawing 27 Amperes from a 12 Volt supply.

In my lab we have set up a CIM with thermocouples on the case and on the brush holder (inside). We will run several cycles at the specified service life load point and record temperatures to make a plot vs. time. When I have that data I will post it. But even without temperature rise measurements, many of us who have run CIMs for driver training while draining several batteries can attest that they can get much hotter than we ever see during FRC match play.